IRM in review Flashcards
Recap of Introductory research methods
What are the 2 types of data analysis?
Quantitative methods & Qualitative methods
Testing theories using numbers
Quantitative methods
Testing theories using behavior/ language
Qualitative methods
The research process
data-Initial observation (research question)-generate theory-generate hypotheses-identify variables-collect data and test theory-measure variables-analyse data-graph data-fit to model
What are the 2 types of measurement
Categorical & Continuous
Entities are divided into distinct categories
Categorical
Entities get a distinct score
Continuous
Discrepancy between the actual value we’re trying to measure, and the number we use to represent the value
Measurement error
Approaches to design
Non-experimental research, Experimental research,
Observe what naturally goes on in the world without directly interfering with it.
Non-experimental research
One or more variable is systematically manipulated to see their effect (alone or in combination with) on an outcome variable
Experimental research
Allows us to make statements about cause and effect
Experimental research
What are the 2 experimental designs?
Independent & Repeated-mesures
Between-group/subjects is what experimental design?
Independent
Within-subjects is what kind of experimental design?
Repeated measures.
Different entities in experimental conditions
Between-group/ subjects.
The same entities take part in all experimental conditions
Within-subjects
First two steps of data ana?
Explore and describe
mean
average score in your data
The middle score of an ordered set
Median
most frequent score
Mode
What is the typical score called?
Central tendency
symmetrical distribution is when…
The mean, mode and median are roughly the same.
skewed distribution is when…
mean, median and mode are different
what are the different modes you can have in the same variable.
bimodal and multimodal
A graph plotting values of observations on the horizontal axis, with a bar showing how many times each value occurred in the data set
Histogram
normal distribution is
bell-shaped and symmetrical around the centre
properties of frequency distribution
skew and kurtosis
symmetry of distri bution
skew
Frequent scores are clustered at lower end with the tail pointing towards high values ie tail point to positive
Positive skew
Frequent scores clustered at higher end with tail pointing toward low values, ie tail points to negative.
Negative skew
Degree to which scores cluster in the tails.
Kurtosis
____means there is a positve kutosis, ie >0 distribution is too peaked.
Leptokurtic
____means there is negative kurtosis, ie <0 distribution is too flat
Platykurtic
when we desire a normal distribution
Normality
what are the two types of variance
parametric and categorical
Standard deviation, range
Parametric
how to check for normality of a histogram
is it bell-shaped, is the distribution symmetrical?
standard deviation is nonsense so use a ratio
Categorical
how to check normality of mean, median and mode?
are they roughly equal
how to check normality of trimmed mean
is the mean similar if you drop the top and bottom 5% of scores?
how to check for normality of skewness and kurtosis
when divided by standard error is it =0
how to check for normality
Histogram, mean median, mode, trimmed mean, skewness and kurtosis
outliers.
how do I check for outliers.
box plot
how to screen multiple variables
homoscedasticity and heteroscedasticity
___are a form of standardization
Z-score
The most frequently occurring score in a variable is the….
Mode
A histogram shows that data in your variable has two “peaks’ rather than a single bell-shape. This is indication of a…
bimodal distribution
A histogram shows that data in your variable is bell-shaped and symmetrical around the center. this is indication of a….
Normal distribution.
When screening the relationship between two variable, scores in a scatterplot tend to be related similarly at low and high values. this is a sign of…
Homoscedasticity
The relationship between two variables seesm to differ at different points in the scatterplot. (e.g) this is a sign of…
heteroscedasticity
A histogram shows that data in your variable is mostly clustered at the high end of values. This is indication of a…
Negative skew
Most of your scores are not dramatically higher or lower than the mean. This would be indicative of…
low standard deviation
A histogram shows that data in your variable has five ‘peaks’ rather than a single bell-shape. This is indication of a…
multimodal distribution
Values vary quite dramatically either side of the mean. This would be indicative of…
high standard deviation
A histogram shows that data in your variable is mostly clustered at the low end of values. This is indication of a…
positive skew
measuring the extent to which two variables are related and measures the pattern of responses actoss two variables.
Correlation
The middle score in an ordered set of data is the…
median
Assesses the linear relationship between continuous variables.
Correlation.
Assumptions of correlation
Linerarity, normality, continuous variables, homoscedasticity.
Possibilty of a third variable.
Tertium quid
Measures the relationship between two variables, controlling for the effect that a third variable has on them
Partial correlation
Measures the relationship between two variables controlling for the effect that a third variable has on only one of the others.
semi-partial correlation.
Its important to ____ _____ ____ because * Reports the magnitude
* There may be a relationship but it might not mean much.
measuring affect size
As the value of one variable increases, the value of the other variable also increases.
Positive correlation.
As the value of on variable increases the value of the other variable decreases
Negative correlation.
strength of the correlation is the R value not the P value. T or F
True
When looking at effect size the R is…
Pearsons product-moment correlation coefficient.
when r is +.1 = small effect, .3 = medium effect and .5 large effect . t or f
true
r2 is….
Coefficient of determination.
by ___ the value of r you get the amount of variance in one variable that is shared by the other
squaring.
test the significance of the difference between means
t-tests
compares two means based on related data e.g. data from the same people measured at different times or data from matched samples.
dependent or paired samples t-tests
compares two means based on independent data e.g. data from different groups of people
Independent-samples t-test.
what are the assumptions for t-tests
normally distributed scored on the DV and variability of scores on the DV is similar for the two categories of IV